Hitachi’s Application of AI and Machine learning

By Hubert Yoshida posted 07-09-2019 00:00



Bernard Marr published an article about The Amazing Ways Hitachi Uses Artificial Intelligence And Machine Learning in the June 14 issue of Forbes. In that article he talks about Hitachi AI Technology/H which was announced in 2015. This AI technology can be applied to many applications rather than just built for a specific application. H supports a wide range of applications and can generate hypotheses from the data itself and select the best options given to it by humans. In the article is a link to a video that shows how a robot was able to teach itself how to swing. “As you can see in the video, the process started with inconsistent movement that doesn't create an effective swinging motion. Over a short amount of time, H tested hypotheses, learned from data, and altered the robot's motion to ultimately create a swinging motion that is more effective than humans produce.”

So far, H has been applied in 57 projects in 14 areas from a warehouse management system to retail sales, call centers, finance, water plants, and more. Bernard goes on to describe research in the use of AI to fight cyber bullying in Japanese schools, urine screening for cancer, food waste, shipping, and finance.

Since Technology H learns from the data it avoids the biases that could be interjected by a model that is developed by a human. It does not require a human to set up a hypothesis in advance, and it can grow according to the situation. 

Technology H was recently adopted by the  The Japan Exchange Regulation ("JPX-R") and Tokyo Stock Exchange, Inc. ("TSE") who decided to take the groundbreaking move of applying artificial intelligence (AI) to market surveillance operations. In conventional market surveillance operations, surveillance personnel conduct preliminary investigations based on preset patterns and analyze the trading situation surrounding such orders. The tests revealed that AI was able to learn on its own and was highly accurate in determining the possibility of unfair trading. This AI initiative facilitates even more in-depth and detailed investigations and is expected to improve market monitoring functions, protecting the fairness and credibility of the TSE market.

Hitachi continues to develop AI, Machine Learning and Deep Learning applications to power good.

For more information on Hitachi AI Technology H see my previous blog post.